BOTANY OF THE LACCADIVES. 469 



plant from Penang, yet Mr. Baker places the Indian plant in 

 0. obtusifolia and regards the Penang one as a variety of 

 C. ensiformis. 



It is therefore better in the meantime to consider C. turgida 

 Grah. to be a plant specifically distinct from C. virosa (the wild form 

 of C. ensiformis), as well as from C. obtusifolia. 



46. Canavalia obtusifolia DC: Baker in Hook, f., Flor. 

 Brit. Ind.j ii., 196 (syn. Doliclios rotundifolius Roxb. excL); 

 Clegh., Madr. Journ. (n. s.), i., t. 4. Doliclios ohcordatus Roxb., 

 Flor. Ind., iii., 303. Probably =C. lineafa DC, Prodr., ii., 404. 



Minikoi ; on sandy beach, Fleming ! 



A littoral species cosmopolitan on tropical shores. 



It is interesting to and on the same island examples of both these 

 sea-coast Canavalias. The specimens of C. turgida are both in 

 flower and with fruit, those of C. obtusifolia are in flower only, but 

 are exactly like the Madras ones (in Herb. Calcutta) of Wallich 

 (Cat. n. 5532), of Wight and of Gamble. They are well dis- 

 tinguished, as Mr. Baker indicates, by the racemes in C. obtusifolia 

 being much the fewer-flowered of the two. But the accuracy of 

 the nomenclature is extremely doubtful, for Canavalia obtusifolia 

 DC (Prodr.y ii., 404) is the exact equivalent of Doliclios obtusifolius 

 Lamk., {Diet., ii., 295), which in turn is, according to Lamarck 

 himself, the plant figured by Rheede [Hort. Malabar., viii., t. 43). 

 It is moreover the equivalent of Doliclios rotundifolius, Vahl (Symb. 

 ii., 81)) of which plant De Candolle himself saw a fruiting specimen. 

 Roxburgh identified the plant described by Vahl with that figured 

 by Rheede, It seems therefore clear that Rheede's Katu-Tsjandi, 

 Lamarck's Doliclios obtusifolius, Vahl's and Roxburgh's Doliclios 

 rotundifolius, De OandoUe's Canavalia obtusifolia and Graham's 

 Canavalia turgida are one and the same sea-coast species , which 

 species is entitled to the name Canavalia obtusifolia. On the other 

 hand, it seems clear from the specimens in Calcutta Herbarium that 

 the plant common on the Madras coast figured by Oleghorn, and 

 the Chinese plant cultivated in the Calcutta Botanic Garden de- 

 scribed and figured {Icon. Ined., xx., 136) by Roxburgh as Doliclios 

 ohcordatus, are specifically identical ; their pods, as figured by 

 Roxburgh and Cleghorn, agree well with the pods of Canavalia lineata 

 62 



